csr and identification of stakeholders

As a management problem theidentification of stakeholders is not easilysolved. It comprises a modelling and anormative issue, which need to be solvedin connection with each other. Instakeholder literature knowledge can befound, e.g. on various stakeholdercategorizations, that could be useful forthe modelling issue. However, thenormative issue remains unresolved.Additionally, the modelling of the so-called stakeholder category ‘theaffected’ further complicates this issue. Nevertheless, from a normativeperspective, this group holds justifiedinterests in aspects of organizationalactivity and its members are, for thatreason, legitimate stakeholders. In this article it is explored to what extentcritical systems heuristics can help inresolving the managerial problem ofidentifying stakeholders. Critical systemsheuristics is a modelling methodology inwhich the normative aspect of modelling

is crucial. Using the distinction between‘the involved’ and ‘the affected’, a varietyof boundary judgments are discussed.Special attention is given to the so-called‘witness’ as a representative of theaffected. Copyright © 2003 John Wiley &Sons, Ltd and ERP Environment.

Received 16 June 2002Revised 14 January 2003Accepted 18 April 2003

INTRODUCTION

Organizations that consider a strategy of corporate social responsibility (CSR)have to address the question ‘to whom

are we responsible?’. ‘To stakeholders’ is thecommon answer to this question, which meansthat, to manage a CSR strategy, the identifica-tion of stakeholders is crucial. To a certainextent, management of CSR has become stake-holder management (Donaldson and Preston,1995).

As a management problem the stakeholderidentification is not easily solved: it comprises,at least, a modelling and a normative issue.The modelling issue refers to questions such as‘Who are our stakeholders?’ or ‘To what extentCopyright © 2003 John Wiley & Sons, Ltd and ERP Environment.

CORPORATE SOCIALRESPONSIBILITY AND THEIDENTIFICATION OF STAKEHOLDERS

Janita F. J. Vos*

University of Groningen, The Netherlands

Corporate Social Responsibility and Environmental ManagementCorp. Soc. Responsib. Environ. Mgmt 10, 141–152 (2003)Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/csr.039

* Correspondence to: Dr J.F.J. Vos, Faculty of Management andOrganization, University of Groningen, P.O. Box 800, 9700 AVGroningen, The Netherlands. E-mail: [email protected]

is it possible to draw the line between stake-holders and non-stakeholders?’.

The normative issue refers to the managerialimplications. Relevant questions are ‘Whatstakeholders do we take into account?’ or ‘Towhat stakeholders are we willing to listen?’.Presumably, this category of stakeholders hasthe capacity to influence managerial and/ororganizational behaviour.

Looking at CSR literature, in a continuing‘stakeholder debate’, these two issues aremingled (cf. Donaldson and Preston, 1995).Much has been written about the concep-tualization of the stakeholder notion (see e.g.Freeman, 1984; Wood, 1991; Clarkson, 1995,Mitchell et al., 1997; Agle et al., 1999; Carroll,1999) that could help in resolving the model-ling issue. However, it is insufficient for thenormative issue. Although the question ‘Whatstakeholders deserve management attention?’has been asked (Mitchell et al., 1997; Agle et al.,1999), it is interpreted in descriptive terms:‘What stakeholders get management attentionfor what reason?’.

To identify stakeholders, in the end, the nor-mative issue needs to be resolved. However,this cannot be done without addressing themodelling issue as well. Before further speci-fying the main question for this paper, someclarifying notes need to be made concerningthe context of CSR.

CSR is defined as the obligations or duties of anorganization to a specific system of stakeholders. Indefining the CSR concept a number of consid-erations have played a role. These considera-tions further clarify the normative perspectivein this paper as well.

First, responsibility is described as ‘having aduty, an obligation’. In this sense, followingTakala and Pallab (2000, p. 111), responsibilitydesignates something still to be done as part of an established role system. The aspect oftime is important here. An organization is not only responsible for past performances –the notion of accountability is relevant then –but it also has a duty in relation to futureactions.

Second, a stakeholder is commonly seen asan individual or a group (see e.g. Freeman,1984). However, it must be emphasized thatthere are relationships between stakeholders.For example, coalitions of stakeholders arelikely to have more influence than a stake-holder alone. Mitchell et al. (1997) describe howdependent Alaskan citizens became, in theirterms, definitive stakeholders of Exxon byacquiring a powerful ally in government.Therefore, from a managerial perspective, it is important to consider also the systemicattributes of stakeholders and refer to a systemof stakeholders. Furthermore, stakeholders arenot objective entities. Using systemic termsemphasizes the need for modelling activities instakeholder management.

Third, the organization is treated here as an entity that is responsible for its activities.Although this is usual in the literature on CSRand business ethics (Takala and Pallab, 2000, p. 112), it has to be acknowledged that this perspective is only possible when humanactivity is involved (Wempe, 1998). In thispaper, CSR is dealt with from a managerialpoint of view, which means, for instance, thatcorporate responsibility leads to managerialresponsibility.

Finally, stressing that a system of stake-holders has to be modelled already involves an element of choice. Asking the question ‘Towhat stakeholders are we willing to listen?’further underlines the element of managerialchoice. It must be emphasized that, in the end,the implication of a stakeholder perspective isa normative one: managers should acknowl-edge the validity of diverse stakeholder interests and should attempt to respond (cf.Donaldson and Preston, 1995, p. 87).

In this paper the problem of stakeholderidentification will be handled from the per-spective of critical systems heuristics (CSH),which has been developed by Werner Ulrich(1983, 1988). CSH is a system approach, whichoffers a variety of boundary judgments.Leaving a more detailed explanation for thefollowing sections, the essence of these bound-

J. F. J. VOS

Copyright © 2003 John Wiley & Sons, Ltd and ERP Environment Corp. Soc. Responsib. Environ. Mgmt 10, 141–152 (2003)

142

ary judgments is that they need to be made ina normative way, which is considered impor-tant in the context of corporate social respon-sibility. This leads to the following question forthis paper: to what extent can critical systemsheuristics help in resolving the managerial problemof identifying stakeholders?

The argument in the paper has been struc-tured along the following lines. First, the stake-holder ‘debate’ in the CSR literature will bediscussed in terms of its consequences for thechoice of stakeholders. Second, the identifica-tion of stakeholders will be discussed as amodelling problem. For that purpose a fewsystem concepts will be explained. Third, onthe basis of this and in order to incorporate the normative aspect of modelling, the line of argument takes a step to critical systemsheuristics. Using the distinction between ‘theinvolved’ and ‘the affected’, the argument goesforward to a number of boundary judgmentsfor modelling a system of stakeholders.

The paper concludes with a review of criti-cal system heuristics as an identification instru-ment for stakeholders in the context ofcorporate and social responsibility.

THE STAKEHOLDER DEBATE IN CSR LITERATURE

This section gives a limited discussion on the stakeholder concept in CSR literature. Themain purpose is to specify its consequences forthe managerial choice of stakeholders.

If we can speak of a ‘stakeholder debate’ inCSR literature at all, it is revealed by means oftwo related questions. The first question is‘What are the responsibilities of an organiza-tion?’ and the second question is ‘To whom isthe organization responsible?’.

Dealing with the first question leads to thenormative content of CSR literature. Respond-ing to, and contradicting, Friedman’s (1997/1970) statement that the social responsibility of business is to increase its profits has playeda crucial role in the normative development

of the CSR literature (see e.g. Ackerman andBauer, 1976; Goodpaster, 1998/1991; Carroll,1999). The main conclusion is, using a varietyof argumentations (see e.g. Freeman, 1994),that there must be attributed more than economic responsibilities to organizations,such as economic, legal, ethical and discre-tionary (or philanthropic) responsibilities(Carroll, 1999). In short, stockholder domi-nance is widely seen as untenable, although, as will be shown subsequently, not by everyone un-problematically (e.g. Goodpaster,1998/1991). In any case, given a variety oforganizational responsibilities, the secondquestion, ‘To whom are we responsible?’, isself-evident.

This leads, directly, to the stakeholder per-spective on organizations. ‘The organization isresponsible to its stakeholders’ is the commonanswer to that question. A straightforwardway of dealing with the stakeholder identifica-tion problem is to make a list of all the keyactors in and around an organization. This maylead to a host of parties: stockholders and othercreditors, consumers, suppliers, governments,employees, management, trade unions, socialcommunities etc. This list demonstrates thedifficulties of such an undertaking. When is the list complete? What makes a party a keyactor? Is it possible or relevant to categorize the various parties in terms of their role orfunction for the organization? Various ways of defining the stakeholder concept result indiverse answers.

Freeman (1984, p. 46) gives a very broad definition: ‘a stakeholder in an organization is(by definition) any group or individual whocan affect or is affected by the achievement ofthe organization’s objectives’. This definition iswidely acknowledged because of its ‘land-mark’ (cf. Wood, 1991; Clarkson, 1995) positionin stakeholder theory. It has been numerouslycited, usually as a starting point to give a narrower view on stakeholders, in which categorizations, different from the distinctionbetween ‘can affect’ and ‘affected’, aredescribed.

CSR AND IDENTIFICATION OF STAKEHOLDERS


143

These categorizations are thoroughlyanalysed by Mitchell et al. (1997). They arguethat the narrower views attempt to definegroups of stakeholders in terms of their directrelevance to the organization’s core interests ortheir necessity for its survival. Clarkson (1995),for instance, makes a distinction between pri-mary and secondary stakeholders. A primarystakeholder group is essential for the survivalof the organization, which is not the case for asecondary stakeholder group.

Elsewhere, Clarkson (in Mitchell et al., 1997)stresses the importance of risk. Without theelement of risk there is no stake. A stake is, inthis sense, something that can be lost. Conse-quently, a stakeholder is a risk-bearer. Fromthis viewpoint the distinction is made betweenvoluntary and involuntary stakeholders. ‘Vol-untary stakeholders bear some form of risk as a result of having invested some form ofcapital, human or financial, something ofvalue, in the firm. In-voluntary stakeholdersare placed at risk as a result of a firm’s activi-ties’ (Mitchell et al., 1997).

Although stockholder dominance is seen asuntenable, to some extent the position of thestockholder is still a point of discussion. Does the stockholder or shareowner have aspecial status among stakeholders? In the so-called ‘Principles of stakeholder management’(Website Redefining the Corporation, 2000),the notion of risk has been used to point outthat, although shareowners may have a spe-cial status, it is not because of higher risks. Onthe contrary, the risks of e.g. employees or cus-tomers may be higher. Shareowners deservetheir special position, according to this state-ment, because of the fact that they have agreed‘that their potential gain or loss from theirinvolvement with the corporation is deter-mined as a residual: it depends upon what isleft over after all other stakeholder claimantshave received their specified distributions’.

Goodpaster (1998/1991) stresses the specialstatus of the stockholder as well. He states that,because of the property rights, the stockholderhas a ‘fiduciary’ relationship with the corpora-

tion. Management fulfils a fiduciary duty to the stockholder, in which trust is crucial. Thismeans that management has the duty to keepthe profit-maximizing promise. In other words,Goodpaster makes a distinction between fidu-ciary and non-fiduciary stakeholders. He posesthe question of whether a multi-fiduciarystakeholder orientation would be desirable. Inother words, could there be ‘fiduciary’ or trust-like relationships with other stakeholders aswell?

Goodpaster rejects this multi-fiduciary position because management will then becaught in what he sees as a stakeholder paradox:‘Management must face resistance from thosewho believe that a strategic orientation (i.e.stockholder dominance) is the only legitimateone for business to adopt’ (p. 113). The paradoxlies in the fact that ‘there is an ethical problemwhichever approach management takes’.Preferably, management should not bear additional fiduciary relationships to thirdparties (i.e. non-stockholder stakeholders), butacknowledge moral obligations directly. Ofcourse there are obligations to other parties.These obligations are moral obligations owedby organizations to those whose freedom and well-being is affected by their activities.These obligations are not instrumental, con-tingent or indirect, but direct or ‘categorical’ (p. 117). It can be concluded that, althoughGoodpaster is a representative of a narrowerview on stakeholders, his analysis of the managerial consequences is consistent with abroader view.

It can be argued that the broad view onstakeholders corresponds with Freeman’s definition. According to Mitchell et al. (1997)this view is based on ‘the empirical reality thatcompanies can indeed be vitally affected by, orcan vitally affect, almost anyone’ (p. 857).

For stakeholder management there are twoperspectives for dealing with this problem, i.e.the firm-centred or instrumental perspectiveand the system-centred or social responsibilityperspective. The firm-centred or instrumentalperspective means that managers want to

J. F. J. VOS


144

know all their stakeholders for firm-centredpurposes, such as economic prosperity,damage control or even image building. It isadvisable to include the affected, because thisgroup can potentially affect the organization’sachievements.

Second, the system-centred or social re-sponsibility perspective corresponds withGoodpaster’s normative position that the organization owes moral obligations to thosewhose freedom and well-being is affected bytheir activities. Whichever position is taken,both perspectives require extensive knowledgeabout actual or potential stakeholders of theorganization.

How do managers deal with this complexproblem? In other words, how do they choosetheir stakeholders and prioritize between com-peting stakeholder claims? Mitchell et al. (1997)and, in a follow-up article, Agle et al. (1999) tryto answer this question by developing the‘stakeholder salience’ model. Stakeholdersalience is defined as ‘the degree to whichmanagers give priority to competing stake-holders claims’ (Mitchell et al., 1997, p. 854).Their claim is that stakeholder salience is positively related to three key stakeholderattributes, i.e. power, legitimacy and urgency,which management believes to be present.

Without discussing the model and the threeattributes in detail, their line of reasoning canbe summarized as follows. Managers perceivea variety of stakeholder groups. They give ahigh priority to a stakeholder if they believethat this stakeholder has a legitimate claim,

which calls for immediate action (i.e. urgent),and possesses the power to influence the orga-nization’s activities. This stakeholder, who isbelieved to possess all three attributes, is calleda definitive stakeholder (category 7 in Figure1). Likewise, a classification of seven stake-holder groups is developed, depending on thepresence of one, two or three attributes invarying combinations.

It goes beyond the scope of this article todiscuss this in more detail. What matters hereis that they have developed a model, whichexplains why managers give attention to whatstakeholders and how they prioritize compet-ing claims.

At this point, two conclusions will be drawn.First, various categorizations of stakeholderscan be found in the literature. In particular, thestudy of Mitchell et al. (1997) has contributedto the analysis of these categorizations. In theintroduction to this paper it was stated that the problem of stakeholder identification com-prises a modelling and a normative issue. Themodelling activity, which will be furtherexplained in the next sections, is about pro-jecting categories on the real world. Therefore,this knowledge about categorizations could beused with regard to the problem of stakeholdermodelling. Leaving aside the question ‘Towhat extent are these categorizations adequatefor modelling purposes?’, it cannot resolve thenormative issue. Once more referring to thesalience model, this study may explain whymanagers give attention to what stakeholders.However, to what extent these management



145

Figure 1. Qualitative classes of stakeholders (adapted from Mitchell et al., 1997, pp. 872 ff ). Reproduced with permission

decisions are justifiable from a normativestandpoint is another matter.

Second, although the distinction between‘affected’ and ‘can affect’ may be widely seenas insufficient for stakeholder identification, itis an important one, especially in the context ofcorporate social responsibility. With respect to this, following Goodpaster (1998/1991),organizations owe obligations to those whosefreedom and well-being is affected by theiractivities. In other words, from a normativeperspective, this group possesses justifiedinterests in aspects of organizational activityand its members are, for that reason, legitimatestakeholders. Acknowledging this, however,further complicates the stakeholder identifica-tion problem. This leads to questions such as‘To what extent and in what way is it possibleto identify the affected? and ‘How far does the inclusion of – potentially – affected stake-holders have to go?’.

With these conclusions and the resultingquestions in mind, we move on to criticalsystems heuristics.

SOME PRINCIPLES OF CRITICALSYSTEMS HEURISTICS

The ideas of critical systems heuristics (Ulrich,1983, 1988; Flood and Jackson, 1991) have beenused here to explore to what extent CSH canhelp resolving the stakeholder identificationproblem, especially with regard to the categoryof the affected. For this purpose, this sectionand the following one explain some principlesand concepts from CSH. Evidently, this willlead to a limited discussion of CSH. The dis-cussion is confined to what is considered rele-vant for modelling and choosing stakeholders.

CSH belongs to the tradition of systemstheory. However, as will be explained below, itoccupies a special position in this tradition.There are many system approaches such as softsystems methodology, system dynamics or theconceptual systems approach. Every approach

has its own characteristics (see e.g. Flood andJackson, 1991; de Leeuw and Volberda, 1996),but they share a strong tradition in method-ologies for modelling and designing systems.These methodologies use a variety of centralconcepts, i.e. system, element, boundary andenvironment. Basically, a system consists of anumber of elements and the relationshipsbetween those elements. Whenever a systemsperspective is applied, decisions must be madethat define the system concerned. It is impor-tant to note that these decisions are determinedby the standpoint or purposes of the researcheror designer.

The most fundamental modelling decisionbounds the system from its environment. Withregard to the stakeholder issue, this could berephrased in bounding the system of stake-holders (note the definition of CSR given in the introduction) from the environment of non-stakeholders.

To a certain extent Ulrich’s ideas can beviewed as a modelling methodology and assuch they rightly belong to the systems tradi-tion. Ulrich develops a number of so-calledboundary judgements by critically reflectingon systems theory in general, and more specif-ically on modelling activity. He advocates a so-called critically normative systems approach(Ulrich, 1988). What matters here is that themodelling activity is acknowledged as a nor-mative activity, which is in need of criticalreflection. A model of a system is adequatewhen it makes its normative content explicit.Before dealing with Ulrich’s boundary judg-ments in more detail in the next section, someprinciples and general notions of CSH will beexplained.

There are four closely intertwined principlesthat guide the practice of CSH (Flood andJackson, 1991, p. 202), i.e. the idea of ‘humanintentionality’, the ‘systems idea’, the ‘moralidea’ and the ‘guarantor idea’. In making plansfor social reality, the ‘planner’ (i.e. thedesigner) has to deal with human intentionsand purposes. Plans – or proposals for design

J. F. J. VOS


146

– have a meaning for individuals. For thatreason, social systems have to be designed tobecome purposeful. A purposeful system isable to produce knowledge that is relevant topurposes and is able to encourage debateabout purposes. In general, according to Floodand Jackson (1991), CSH is about the designand assessment of purposeful systems. Withregard to this, it is important that it is not onlyabout ‘how to do things’, but that it helps us to decide what we ‘ought to do’ in order toimprove reality.

The systems, the moral and the guarantoridea are three ‘quasi-transcendental’ ideas,which are based on the philosophy of Kant andshould be used as critical standards in theprocess of making plans for social reality. Basically, with the systems idea Ulrich stressesin a philosophical way what is common inevery system approach. A system is a model (of a part) of social reality, which is always a limited projection of that part of social reality. With the systems idea Ulrich empha-sizes the necessity to consider this ‘lack of comprehensiveness’.

Through the moral idea Ulrich stresses that the planner should aim to improve humanconditions by means of his/her plans. Thismeans that the planner must be aware of themoral implications of the plan. As will beexplained in the following section, referring tothe so-called ‘involved’ and ‘affected’ of thesocial system reveals this awareness of theplanner.

The meaning of the guarantor idea is that,although there is no guarantee that planningleads to improvement, the planner should seekto incorporate as many sources of guarantee aspossible. This leads to seeking opinions by theplanner from as many experts and stakeholdergroups as possible (Flood and Jackson, 1991, p. 204).

In the following section these experts andstakeholder groups are interpreted in terms ofthe involved and affected in connection withthe necessary boundary judgments.

MODELLING A SYSTEM OFSTAKEHOLDERS: THE INVOLVEDAND THE AFFECTED

The main conclusion so far is that the planner,during his designing activities, must refer to avariety of groups. At this point it is relevant toreturn to stakeholder management in the CSRpractice. After that, we will proceed with themodelling of the involved and affected as specifications of stakeholder groups.

In the previous section it is explained thatthe planner fulfils the role of a designer inorder to improve social reality. This means thatthe notion of planning must be interpretedvery broadly; planning is understood as socialsystems design. The concepts of ‘planning’ and‘planner’, as used by Ulrich, are used here interms of ‘stakeholder management’ and ‘stake-holder manager’.

According to Ulrich (1983, p. 24), planningis, or should be, a public activity if the groupof involved planners is not identical with theaffected citizens. This position is very inter-esting within the context of stakeholder management. It means that stakeholder management should be a public activity if the group of involved beneficiaries of organi-zational activities is not identical with the affected non-beneficiaries.

Ulrich explains the boundary judgements byspecifying the various groups of social actorsto whom the planner must refer. With regardto the notions of social actor and planner someremarks must be made.

The term social actor is interpreted here as astakeholder. Actor refers to a social role ratherthan individuals. This means that these roleshave to be specified in a concrete case in orderto decide what individuals or groups of indi-viduals are representatives of what roles. The‘planner’ must be seen as a social role as well.As said previously, in relation to stakeholdermanagement he fulfils the role of stakeholdermanager. Furthermore, in order to guaranteethe normative content of the modelling prac-



147

tice, the various roles and their concerns in rela-tion to the social system have to be specified.

Ulrich (p. 248) acknowledges two reasonsanyone can claim belonging to the system: (i)because he is actually or potentially affected bythe outcome of the system and (ii) because hehas some kind of resource (expertise, politicalor financial, etc.) to contribute to the system,i.e. because he is involved. This leads to thetwo basic boundary judgments. The first judg-ment bounds the total system of the involvedand the affected from the environment, that isfrom the environment of non-stakeholders.The second one makes the distinction betweenthe involved and the affected. The termaffected is restricted to the group that isaffected, but not involved. This means thatthose individuals or groups who are involvedand affected are considered as belonging to thegroup of the involved.

Let us consider the two basic groups in more detail and begin with the involved. Onthe basis of three sources of influence Ulrichdivides the involved into three sub-groups.

(i) Sources of motivation: whose purposes(values, interests) are being (ought to be)served? This leads to the group ‘client’.

(ii) Sources of control: who has (ought tohave) the power to decide? This leads tothe group ‘decision maker’.

(iii) Sources of expertise: who has (ought tohave) the necessary expertise? This leadsto the group ‘planner’.

Three notes are relevant. First, concepts or categories such as ‘involved’, ‘affected’ or‘planner’ are projected on aspects of the realworld. This is what modelling is about. Wedecide to call a certain group with certain char-acteristics ‘client’ or ‘planner’. What may beconfusing here is that the application of CSH isnot limited to the practice of organizationaldesign or change etc., but can also be used forpractices such as policy development or com-munity planning. In organization theory theterms client or planner usually have a morerestricted meaning.

Second, every group can be considered aprototypical role and comprises a specific caseof a variety of individuals, who may belong tomore than one group.

Third, every question of which the answerleads to a boundary judgement should notonly be phrased in the is mode, but also in theought mode. In this way, according to Floodand Jackson (1991), the normative content ofthe system design is best seen.

The second basic group, the affected, is moredifficult to specify. This group does not have awell defined contribution, but they endure theunwanted side effects or pay some costs thatare not endured by those who benefit from thesystem. As noted before, especially concerningthe affected the problem of identifying stake-holders gets through. Potentially, this group isvery large and it is not clear where the bound-aries of this group lie.

According to Ulrich (1983, p. 251) theaffected groups can rarely be bounded,because knowledge about the potential sideeffects and long-term risks is usually insuffi-cient (see e.g. the recent discussions about theunknown effects of changing the biologicalgenes of our food). Because of this identi-fication problem Ulrich concludes that ‘theaffected’ can only be bounded by means of arepresentation. It is crucial that not theinvolved, i.e. the planner, but the affectedthemselves will determine who is to representthem.

At this point, the question rises of whetherwe have made any progress in resolving theproblem of identifying this type of stakeholder.Indeed, as we look at the definition by Freeman(1984) of a stakeholder as ‘any group or indi-vidual who can affect or is affected by theachievement of the organization’s objectives’,there is a striking similarity with the definitionof the involved (i.e. can affect) and the affected(i.e. is affected).

What makes Ulrich’s (1983, p. 252) distinc-tion noteworthy is that he comes up with a rolethat will argue the case of the affected; theessence of this role is that of a witness. Ulrich

J. F. J. VOS


148

states ‘. . . Rather than pretending that we canadequately grasp such issues by means of adetailed list of heuristically necessary cate-gories, I suggest that we limit the claim forheuristic necessity to the one essential cate-gory of the “witness” . . . the planner cannotadequately trace the normative content of alternative boundary judgements . . . withoutreferring to some social actors playing the roleof witness . . .’.

So far we have four categories of socialactors that comprise the system of stake-holders, and should have an input in the orga-nization concerned. The client, the decisionmaker and the planner are directly involved inthe outcome of the organization. The witnessrepresents the affected, which means that hebecomes – indirectly – involved.

In order to find the boundaries of the systemand, more importantly from a normative pointof view, to reveal what the boundaries ought tobe, there are three questions for every groupthat explore their sources of influence. The firstand basic question is rather simple; it refers tothe social role of the group. Who is the client,or who ought to be the client of the system?Likewise, similar questions are posed for theother groups.

The following two questions are group spe-cific and fulfil an auxiliary function in order tofurther determinate the groups. They refer to

‘role specific concerns’ (i.e. the (i) questions) orto the key problems with regard to the deter-mination of that group (i.e. the (ii) questions).Figure 2 summarizes the complete set ofboundary questions.

Although it is assumed that the complete set is useful for modelling and choosing stake-holders, this paper will not examine the com-plete set of questions in more detail (see Ulrich,1983; Flood and Jackson, 1991). For reasons ofspace and because identifying ‘the affected’ ismost problematic, the remainder of this paperfocuses on the so-called witness.

The Witness: a Representative of the Affected

While it is possible to describe the roles of theclient, the decision maker and the planner infunctionalistic terms, that is not the case for thewitness. Witnesses do not contribute resourcesor expertise, nor do their goals motivate orga-nizational activity.

Ulrich (1983, p. 256) reasons that the wit-nesses represent what he calls ‘the crucialsource of legitimation’. By arguing the case ofthe affected towards the involved, they remindthe involved of their moral responsibility forall the practical consequences of organizationalactivity. It is important that this responsibilitygoes beyond the effective functioning of theorganization.



149

Figure 2. A system of stakeholders (adapted from Ulrich, 1983, pp. 252 ff ). Copyright © 1983 John Wiley & Sons Limited. Reproduced with permission

Basically, the different concerns of theinvolved and affected may be fundamentallyconflicting. The essential point is, according toUlrich, that the affected must be given thechance of emancipating themselves from beingtreated merely as means for purposes of othersto ‘an end in themselves’. This is why Ulrichdesignates ‘emancipation’ as the major concernfor the witness. Accordingly, this leads to thekey problem (i.e. question 4b in Figure 2) forthe determination of this group. Conflictingconcerns may be rooted in different world-views (weltanschauung). In order to account forthis fact, the different world-views of theinvolved and the affected must be traced.

Who could be considered witnesses and inwhat way do they fulfil their role? In general,one could think of action groups, pressuregroups and the media when, and solely in thiscase, they argue the case of the affected withregard to a specific issue. A short case studyabout a nuclear reactor will further exemplifythe role of the witness.

Very recently, the nuclear reactor at theDutch town of Petten has been temporarilyshut down for a review of safety procedures.Up to this closing down on 18 February 2002,there was an interesting discussion to be seenbetween various stakeholders with conflictinginterests. The plant at Petten is a so-called highflux reactor. The Joint Research Center, as partof the European Commission, who is theowner, has set up this plant for medical pur-poses. The permission to operate is issued by the Dutch Environment Minister and theNuclear Research and Consultancy Group(NRG) operates the plant. The latter can beconsidered an operating company.

It is worth mentioning that nuclear energy isa very sensitive issue in the Netherlands.Almost all nuclear power has been phased outand only one plant – at Borssele – is still inoperation. Even though the reactor in Pettenexists for medical purposes, there has also beendiscussion about a possible closure of thisreactor. In this discussion, several issues play arole, such as safety, what to do with nuclear

waste and the production of nuclear materialfor cancer research.

In December 2001, ‘a whistle-blower’ fromwithin the NRG made his concerns about the safety procedures public. Leaving aside the preceding and resulting problems in theworking relationships, through the whistle-blower the events developed rapidly. An investigation of the Nuclear Physical Agency, agovernmental supervisory body, points at hair-cracks and procedural irregularities. Whereasit was stressed that the reactor is completelysafe, the environment minister, at that time JanPronk, asked the owner of the facility to tem-porarily close it down. After some discussionbetween the Dutch minister, the EuropeanCommission and the NRG, the NRG compliedwith the urgent request to shut down thereactor. This seems to be, although temporar-ily, the end of story.

However, immediately after the closingdecision, health-care arguments gained impor-tance. There were various press reports thatquoted cancer specialists about the negativeconsequences for cancer patients. Because thereactor produces radio-isotopes for Europeanhospitals, closure could endanger the diagno-sis and therapy of cancer patients. After con-sulting the healthcare minister, the closure waspostponed for two weeks.

This short illustration draws attention tosome of the conflicting interests between stake-holders. It must be stressed that the case is farmore complex than described here; for reasonsof space it is not possible to analyse the casemore thoroughly. The category of the affectedcomprises at least two groups, depending fromthe issue that is under discussion. First, fromthe safety perspective the ‘possibly affected’are Dutch or even European citizens. Althoughthe whistle-blower belongs to the category ofthe involved, he can be viewed a witness forthis type of affected as well.

Second, from the healthcare perspective, theaffected are cancer patients, aware or notaware of their illness. In this matter variouscancer experts and the media fulfil the role of

J. F. J. VOS


150

the witness and cause a commotion about thepossible negative effects of the closure. Even-tually, the closure is reconsidered.

CONCLUSION

This paper has focused on the question ‘Towhat extent can critical system heuristics(CSH) help in resolving the managerialproblem of identifying stakeholders?’ It hasbeen argued that this identification problemcomprises a modelling and a normative issue.The practice of corporate social responsibilitymeans responding to a variety of stakeholderswith different interests and needs. Conse-quently, there are always decisions to be made among conflicting interests and needs.Although CSH is by no means – in the termsof Ulrich – a guarantor for adequately dealingwith those conflicting interests, it positivelyhelps us to become aware of them. Further-more, by answering the boundary questionsthe system of stakeholders can be modelledand this helps to reflect on the implications ofdecisions made. Generally, it can be concludedthat CSH has the potential to deal with themodelling and the normative issue concur-rently. This conclusion, as such, makes itworthwhile to further explore the stakeholderidentifying capabilities in the context of corpo-rate social responsibility.

Undoubtedly, empirical research shouldfurther explore and clarify the practical rele-vance of CSH for stakeholder identification ingeneral, and for identification of the affectedmore specifically. Although this paper hasfocused on the witness as a stakeholder cate-gory, it is assumed that the complete set ofboundary judgments is helpful for identifyingstakeholders, i.e. affected and involved.

After these more general remarks, the fol-lowing conclusions refer to the identification ofthe affected and to the witness as a represen-tative of the affected. First, the distinctionbetween the involved and the affected is con-sidered crucial, not only because the involved

have the possibility to influence organizationalbehaviour and the affected do not have. Thedistinction is particularly important because of the modelling problems with regard to theaffected. Acknowledging these problems alongwith the moral claim that an organization owesobligations to those whose freedom and well-being is affected, the notion of the witness pos-sibly offers a way out. It reduces the modellingproblems, not least because the initiative lieswith those who consider themselves affected.The witness stands up for the interests of a specific group, as some medical experts did forthe cancer patients in the ‘Petten case’. As alsoshown in this case, the affected is not a homo-geneous group for an organization. It is likelythis category comprises different groups withvarying interests, depending on the decision(s)on the agenda.

Second, the witness can be viewed a channelfor the affected to influence the organization,which means that, to some extent, they becomeinvolved. However, there is another side of thepicture. As said before, there are always deci-sions to be made among conflicting interestsand needs. Sometimes it comes down tofinding a responsible balance between twoevils. This is what Wempe (1998) calls the ‘dirtyhands dilemma’. Finding a responsible balancemakes detailed investigation into the nature ofthe dilemma necessary. The notion of a witnessmay offer a means to investigate the nature ofvarious stakeholder claims.

Third, it appears relevant to differentiatebetween witnesses depending on the amountand nature of their influence. There is a differ-ence between a witness who stands up for theinterests of a specific group and a witness whofulfils the role of a ‘mediator’ between theorganization and a group of affected. A witnessmobilizes forces to influence the organization;a mediator tries to settle the differences.

Finally, an important question is ‘When can an individual rightfully be considered a witness?’. An individual is solely a witnesswhen he or she represents a group or individ-ual that is affected by the organization’s activ-



151

Donaldson T, Preston LE. 1995. The stakeholder theory ofthe corporation: concepts, evidence, and implications.Academy of Management Review 20(1): 65–91.

Flood RL, Jackson MC. 1991. Creative Problem Solving.Total Systems Intervention. Wiley: Chichester.

Freeman RE. 1984. Strategic Management: a StakeholderApproach. Pitman–Ballinger: Boston.

Freeman RE. 1994. A stakeholder theory of the moderncorporation. In Ethical Theory and Business 5th edn,Beauchamp TL, Bowie NE (eds). Prentice-Hall: UpperSaddle River, NJ; 66–76.

Friedman M. 1997. The social responsibility of business isto increase its profits. In Ethical Theory and Business 5thedn, Beauchamp TL, Bowie NE (eds). Prentice-Hall:Upper Saddle River, NJ (original work published 1970in New York Times Magazine 13 September).

Goodpaster KE. 1998. Business ethics and stakeholderanalysis. In The Corporation and its Stakeholders. Classicand Contemporary Readings, Clarkson MBE (ed.). University of Toronto Press; 103–123 (original workpublished in 1991).

Mitchell RK, Agle BR, Wood DJ. 1997. Toward a theory ofstakeholder identification and salience: defining theprinciple of who and what really counts. Academy ofManagement Review 22(4): 853–886.

Takala T, Pallab P. 2000. Individual, collective and socialresponsibility of the firm. Business Ethics: a EuropeanReview 9(2): 109–118.

Ulrich W. 1983. Critical Heuristics of Social Planning. A NewApproach to Practical Philosophy. Wiley: Chichester.

Ulrich W. 1988. Systems thinking, systems practice, andpractical philosophy: a program of research. SystemsPractice 1(2): 137–163.

Website Redefining the Corporation. 2000. http://www.mgmt.utoronto.ca/~stake/Principles.htm[31 January 2002].

Wempe J. 1998. Market and Morality. Business Ethics and theDirty and Many Hands Dilemma. Eburon: Delft.

Wood DJ. 1991. Corporate social performance revisited.Academy of Management Review 16(4): 691–719.

BIOGRAPHY

Tel: Dr Janita F. J. Vos is Assistant Professor atthe Faculty of Management and Organization,University of Groningen, P.O. Box 800, 9700 AVGroningen, The Netherlands.Email: [email protected]

J. F. J. VOS


152

ities. It is the role of the witness to argue thatthe claim is justified. It must be emphasizedthat the reverse is not necessarily true. In otherwords, someone who witnesses, e.g. in themedia, an organizational problem or failure isnot necessarily affected by this event nor is arepresentative of the affected. This is why Ulrichstresses that it is only the affected themselveswho can determine who is to represent them.Even so, a proactive attitude from the perspec-tive of stakeholder management seems to beimportant. This leads to questions such as ‘Inwhat way are witnesses determined?’, ‘Whatkind of witness is available or adequate?’ and‘Whom are they representing?’.

ACKNOWLEDGEMENTS

An earlier version of this paper was presented at the 2001Annual Research Conference on Sustainable Develop-ment. The author is grateful for comments received afterthe presentation, and especially from the anonymousreviewer.

REFERENCES

Ackerman RW, Bauer RA. 1976. Corporate Social Respon-siveness, The Modern Dilemma. Reston PublishingCompany: Reston, VA.

Agle BR, Mitchell RK, Sonnenfeld JA. 1999. Who mattersto CEOs? An investigation of stakeholder attributesand salience, corporate performance, and CEO values.Academy of Management Journal 42(5): 507–525.

Carroll AB. 1999. Corporate social responsibility. Evolu-tion of a definitional construct. Business and Society38(3): 268–295.

Clarkson MBE. 1995. A stakeholder framework for ana-lyzing and evaluating corporate social performance.Academy of Management Review 20(1): 92–117.

de Leeuw ACJ, Volberda HW. 1996. On the concept of flexibility: a dual control perspective. Omega, International Journal of Management Science 24(2): 121–139.

csr and identification of stakeholders

Documents